558
INDUSTRIAL A N D ENGINEERING CHEMTSTRY
Vol. 19, No. 5
Some Variables Affecting the Behavior of Limes Used in Causticizing' By J. V. N. Dorr and A. W. Bull THE DORRCOXPANY, NEWYORK,N. Y.
This paper discusses some of the factors involved in the choice of limes for causticizing. Experimental results are given to show that the method of causticizing is of great importance in determining the physical properties of the calcium carbonate produced, and it is pointed out that this in turn will control the size of chemical equipment needed for a causticizing plant with given capacity. The nature of the causticizing reaction is discussed and it is shown that in reactions of this kind the size of particles in the final precipitate may be determined by the size of particles of the lime after slaking.
H E preparation of caustic soda from soda ash and lime is one of the oldest industrial processes and one of the simplest from the standpoint of the chemical reaction involved. Soda ash and calcium hydroxide react to form caustic soda and calcium carbonate in accordance with the equation: NazCOa 4- Ca(0H)z = CaC08 2NaOH The choice of lime for carrying out this reaction has been the subject of considerable study and several papers have been published, including a recommended specification for lime for causticizing, by the Bureau of Standards,2 which states: The standard of composition for quicklime for use in causticizing shall be a content of 85 per cent of available lime. Lime containing more than 3 per cent magnesia or less than 70 per cent of available lime shall be rejected as uneconomical to use. It is recommended that a bonus or a penalty of l l / z per cent of the contract price be added to or deducted from the payment for each 1 per cent of available lime above or below the standard 85 per cent.
T
+
In choosing lime for causticizing it is necessary to consider its physical as well as its chemical properties, because after the causticizing reaction has been completed the caustic soda solution must be separated from the precipitated calcium carbonate and the ease with which this can be done is dependent to some extent on the physical properties of the lime used. The separation of the caustic soda solution from the calcium carbonate mud can be readily accomplished by decantation, preferably with countercurrent washing of the calcium carbonate, in a series of decantation tanks. Soon after Dorr thickeners were first applied to this problem of continuous countercurrent decantation it became evident that the settling rates of the calcium carbonates produced a t different plants varied widely, with a corresponding variation in the size of thickeners required to handle a given tonnage of solids. At f i s t these variations were considered to be due principally to the use of different lipes, but recent experiments indicate clearly that a number of other factors may be of great importance. Factors Influencing Settling Rate
Among the factors which are considered to have an influence on the settling rate of the calcium carbonate sludge after causticizing are the following: Presented under the title "Limes Used in 1 Received March 21, 1927. Causticizing and Some Variables Affecting Their Behavior." Bur. Standards, Circ. 143.
I-Source
of lime
A . Chemical constitution B. Physical nature 11-Method of burning the lime A . Temperature B. Length of the burning period
111-Method of slaking the lime A . Amount of water used B. Temperature during slaking C. Degree of agitation during slaking D. Use of soda ash or caustic soda solutions for slaking IV-Method of causticizing A . Period of agitation B. Violence of agitation C. Temperature D. Excess of lime or soda ash used V-Causticity and concentration of caustic desired VI-Presence of impurities particularly when reburned lime is used Harrop and Forrest3 have shown the influence of some of the factors involved. Stewart and Walmsley' have given additional data, although their complete paper does not appear to have been published. Holmes, Fink,and Mathers' have studied the settling rates of different limes after slaking. Recent papers by Haslam and Hermann6and Haslam, Adams, and Kean7 have shown that the time and temperature of burning have an important influence on the settling rate of the lime after hydration. Whitman and Daviss have studied various methods of hydration. I n considering the question of lime for causticizing it is therefore important that the many variables involved be carefully considered. It is the intention of this paper to discuss the settling behavior of various high-calcium limes and to show the great influence which changes in the method of slaking and causticizing have upon the settling rates. Source of Limes Studied
I n order to compare the behavior of lime from different sources, samples were obtained from eleven lime companies through the cooperation of the National Lime Association. Two of these samples were from Virginia, the other nine being from Pennsylvania, Georgia, New York, Massachusetts, West Virginia, Maryland, Illinois, Maine, and Texas. With two exceptions, these samples were all high-calcium limes ranging from 91.3 to 96.9 per cent calcium oxide by the sugar method, and all were standard products produced by normal kiln operation. Experimental Procedure
Experiments were first made using only one of these limes under different conditions. Later the different limes were directly compared under a standard set of conditions. The apparatus (Figure 1) for carrying out the causticizing reaction and for making the settling-rate tests consisted of a large box with a glass window and a thermostatically controlled heating unit. The 2500-cc. beaker containing the ~ T m sJOURNAL, 16, 362 (1923). 4 J . SOC.Chem. Ind., 43, 251R (1924). 5 Chem. & Met. Eng., 27, 1212 (1922).
THISJOURNAL, 18, 960 (1926). Ibid., 18, 19 (1926). ' I b i d . , 18, 118 (1926).
7
May, 1927
INDUSTRIAL A N D ENGINEERING CHEMISTRY
solution rested on a small heating unit connected with a variable resistance. A propeller could be operated at any desired speed by the variable-speed motor, although unless otherwise specified the speed was 180 r. p. m. Water could be added through the small funnel to compensate for evaporation. It is important that the temperature of the air and of the solution be practically the same during the settling tests, for otherwise convection currents are set up which appreciably affect the settling rate. The heating unit under the beaker was constructed to have a minimum heat capacity, so that when it was cut off just before starting the settling test it cooled quickly and did not cause convection currents. After the causticizing reaction had been completed, the propeller was raised above the solution level and settling tests were made directly in the beaker. The soda ash used was carefully dried and analyses showed it to contain 98.9 per cent Na2C03. I n most of the tests 200 grams of sodium carbonate and 1300 CC. of water were used. After causticizing, the solution was approximately 15" BB. (at 20" C.) in all cases, since evaporation was compensated for by small additions of' water during the test. Effect of Various Conditions on Settling Rate
I n the first tests 118.5 grams of lime, corresponding to 3 per cent excess over that theoretically required to convert the soda ash completely to caustic soda, were slaked in 500 cc. of water initially a t 85" C. and stirred for 15 minutes, after which the soda ash, dissolved in 800 cc. of water, was added and stirring was continued for 30 minutes longer. The average settling rate of the calcium carbonate in three tests under these conditions was 0.16 foot (0.049 meter) per hour. Although this settling rate is very low, even slower rates have been observed in some causticizing plants. SXALLAMOUXTOF WATER FOR S ~ a ~ ~ x ~ - - P r e v i oinus vestigators have shown that the amount of water used in slaking the lime is important, so experiments were made to see whether the settling rate could be increased by changing the method of slaking the lime. Instead of slaking with 500 cc. of water, only 80 cc. were used and the dry hydrate was formed. This was then stirred with 550 cc. of water for 10 minutes to be sure that hydration was completed, after which the soda ash was added in solution. The mixture was stirred for 30 minutes and at the end of this period the settling rate was found to be 1.41 feet (0.43 meter) per hour or 8.8 times the rate found when the lime was slaked in a large excess of water. To be certain that this very large increase was not due to some peculiar characteristic of the particular lime tested, two other limes were compared in the same way. I n one case the settling rate increased from 0.28 to 1.48 feet (0.085 to 0.45 meter) per hour and in the other case from 0.19 to 4.06 feet (0.058to 1.24meters) per hour. The average settling rate for the three limes is increased nearly twelve times by reducing the water used in slaking. EXCESS LIME-with this evidence that other limes showed the same general behavior as the one originally chosen, experiments with the original lime were continued. Tests were made to determine the effect of excess lime on the settling rate. I n each case the lime was slaked just before the test by slowly pouring over it 100 cc. of hot water. After the hydrate thus formed had cooled, it was added to a solution of 200 grams of sodium carbonate in 1350 cc. of water at 85" C. and the mixture was agitated for 30 minutes, the temperature being maintained at 85" C. Before determining the settling rate in the last test, caustic liquor was added to reduce the ratio of solids to liquid to that in the test with 115 grams of lime. Unless the ratio of liquid to
559
solid is kept uniform the settling rates cannot be fairly compared. The data are given in Table I.
CaO
Grams 100 115 130 150
Table I-Effect of Excess Lime PERCENT OF THEORETICALCAUSTICITY SETTLING RATE Per cent Ft./hr. M./hr.
a7 100 113 150
79.9 87.6 92.7 94.1
2.97 2.97 2.82 2.12
0.91 0.91 0.86 0.65
Within reasonable limits it is evident that an excess of lime has only a slight effect on the settling rate. TEMPERATURE A S D TIME O F AGITATION-The effect Of temperature and of the period of agitation was then studied. Portions of the standard lime weighing 125 grams were slaked with 100 cc. of hot water and then added to solutions containing 200 grams of sodium carbonate in 1350 cc. of water. Agitation was continued a t the temperatures shown. The data are given in Table 11. It is evident that the settling rate decreases as the agitation is Continued, and it is also evident that temperatures above 85" C. are of little benefit. Under the conditions of hydration and agitation used in the tests the reaction is nearly a t equilibrium a t the end of an hour. The causticity figures were obtained by titrating a portion of the supernatant liquor, after the test, with normal hydrochloric acid using phenolphthalein and methyl orange. "Causticity" has been calculated using the formula 100(2PP-MO) Per cent causticity = 11.10
where $10 is the number of cubic centimeters of hydrochloric acid required to make the solution neutral to methyl orange
Figure 1-Apparatus
for Causticizing and Settling-Rate Tests
and PP is the number of cubic centimeters required to make the solution neutral to phenolphthalein. Per cent causticity in this case is equivalent to per cent conversion or to t h e percentage of the original sodium carbonate which has been changed to sodium hydroxide. Some writers have used "causticity" as equivalent to per cent caustic or equal to 100 (NaOH) This formula gives lower numerical values NaOH Na2C03. +
than the formula which was used in these tests. LIME SLAKEDIN HYDRATOR-Instead of slaking the lime by pouring water over it, it was decided to duplicate more closely the industrial operation for preparing hydrated lime by using a small hydrator arranged by placing a one-quart
ISDCSTRIA41. --LYD ESGISEERI,L'G CHEJfISTRY
560
metal can on revohiiig roil< 1 m a l l hole in the cover allomed steam to escape during the hydration. The lime was placed in the can, 45 to 50 cc. of cold water were quickly added, the cover wa4 then put on. and the can was placed (111the rolls The slaking reaction began almost iiiimediately xiid was usually practically complete within 2 minutes. However, the can wa< allowed to rotate for 10 riiinutes before the hydrate w a i renioyed. All hydrated lime used in the later te-ts ma- prepared iii thi-. way.
70' C .
I
Comparative Tests on Different Limes Comparative tests n-ere next made on elel-eii different limes. I n each case an amount of lime equivalent to 105 per cent of the quantity theoretically required to convert all the sodium carbonate to sodium hydroxide was slaked in the hydrator with 45 to 50 cc. of water and was then added t o 1350 cc. of soda ash solut.ion containing 200 granis of sodium carbonate. The mixture was agitated for one hour at 95" C.,
of Temperature and Period of Agitation
Table 11-Effect I
1-01. 19, s o . 5
8.53 C.
~
I
0 5 O C.
.\CITATION
Causticity
, __ ~
Settling rate
~
Causticity
Settling r a t e
3.73 2.97 2.34 1.96 1.56 1.33 1.10 1,lR
SLOWADDITIO?;OF SODALlSH-rt n-as thought possible that gradual addition of the soda n.sh might decrcase tlie rate of formation of thr c,alcium carbonate n-itli a consequent increase iii the size of the calcium carbonate crystals and an increase iii 9ettling rate. To test this point. 125 grams of the staiidard lime were hydrated mid added to 650 cc,. of water a t 95" C. Two hundred graiiis of soda ash were dissolved in TOO cc. of hot water aiid the solution was added to the lime auspeiision i n a slow htreaiii over a period of half an hour. the lime suspension being kept at. approximately 100' C. during this period. lgitatioii was continued for a n hour after all the soda ash solution had been added. ;It the end of thii time the settling rate was fouiid to be 1.19 feet (0.36 meter) per hour and thecausticity was 92.5 per cent. Slow adclitioii of sodium carbonate does not a m e a r to he desirable m c e tlie settling rate waq lon ered rather than increased 1% hen uqing this method. HYDR ~TIO?; WITH CAUSTIC Son.\--Another test was made to determine the effect of hydrating the lime with caustic soda solution instead of with water. Standard lime weighing 125 grains was placed in the hydrator and 47 cc. of a 14.8" B6, caustic solution, prepared in one of the previous causticizing tests, were added for hydration. The lime was considerably longer in reaching a maximum teniperature than in the usual tests when water was used. The hydrate was added t o the hot soda solution and the mixture was stirred for 90 minutes. Tests made during this period gave the data shown in Table 111. A
Causticity
1.14 0.91 0.71 0 ,f i l l 0.1s 0.41 0.36 0.33
I t is evident that this method of preparing the hydrate yields an extremely rapid settling carbonate after causticizing. As agitation is continued the settling rate decreases, as previously noted. VARYIKG DEGREE OF AGITATION-TOdetermine the effect of changing the degree of agitation, comparative tests were made with the propeller running a t 180 and a t 420 r. p. In. The lime was hydrated with water as usual and added to the soda ash solution, the temperature being held at 85' C. The data are given in Table IT-. Increasing the degree of agitation decreases the settling rate of the carbonate produced, but increases the rate of the cnusticizing reaction.
i
inoc
Causticity
c. Settling r a t e
92 0
$11 Ill
s
the propeller riiiiiiing at 180 r. p. in. Settliiip :tiid wuaticity testh were iiiatle after 30 iiiiiiutes and agaiii d t e r oiie h(111r. The data are gil-en iii Table T-. -Although there is a considerable variation iii tlic settling rate of the different limes, their behavior ih iiiore uiiiioriii than might be expected. It mill be noted that, ilk giieral, higher causticities are obtained with the alon.er wttlkig carbonates. This seems to be a general rule. I t w a ~also observed that the very rapidly settling carlmiatci hat1 a tendency to leave some fine material in suspension. n.licv:a-: the very slon- settling carbonate, being very bulky. cxrrierl d o w i all the fine material leaying a practically crwtal-clear solution to be overflowed or drawn off. Table IV-Effect
I
Table 111-Lime Hydrated with Caustic Liquor AGITATION CAWSTICITI SETTLING RATE Mcnules Peu cent Ft / h r M /hr 2 3s 7.81 91 ;i 30 2.00 92 8 6 . 5 7 60 1.72 5.64 94 4 90
Settling r a t e
1
AGITATIOK I
i
of Varying the Degree of Agitation
isn R . P. M.
1
Causticity
420
Settling r a t e
Pur c i ' n ~
Ft,/hr.
94.7
2.s0
A\1./12~. 0.i6
Causticity
Per cent 95.0
R. P
31.
Settling r a t e 1;/ V i r . i.i;i'
-11
'AY.
051
60 90
izn
1,IME S O .
1 3
8
9 10
11 ._
1
30 MISWTES'AGITATIOK Causticity
Pprrent 94.i 93.4 94.6 93 4 94.3 91.4 90.5 91 2 9L.3 90.2 ~. ~
Settling r a t e
F!./hr. 2.50 1.96 1.33 1.33 1.56 3.33 2.08 5.63 1.46 5.31
M,/hr. 0.76 0.60
60 MINUTES'. ~ G I T A T I O S Causticity
PPVcent
Settling r a t e Ft./iir.
.11.//7t.
0.41 0.41
0.48 1.01 0.63 1.72 0.44 1.62
Nature of Causticizing Reaction The fact that rapidly settling hydrates yield rapidly settling carbonates after causticizing makes it appear probable that in the causticizing reaction the calcium hydrate particles do not pass into solution to be reprecipitated froni solution, but that the carbonate is formed a t the surface of the hydrate particle and that a gradual replacement of hydrate by carbonate occurs by diffusion in the particle. I t seems quite probable that such replacements are comparatively common in those reactions in which a slightly soluble iiiaterial is converted to a less soluble material and that the phenomenon is of general interest. The writers have been unable to find
yielded much coar,.er aggregate3 of liariuiii sulfate with a correspondingly rapid settling rate. 0 1 1 the other hand, the reaction was much slower ant1 the supernatant liquor after the bulk of the material had settled was decidedly more turbid when the coarser hydrate was used. These re.sults are very similar to the data obtained from causticizing. I t seems probable that similar conditions exist in a iiuinber of industrial processes niid that we should be able to alter the physical characteristics of the filial insoluble material b y modifying the characteristics of one of the factors iri the reaction when the latter is used as a suspension rather thali as a true solution. Hon-ever. v-hen the reaction is accoiiipanied by a comparatively large change in the ~ d i i r i i eof the solids involved. it \ d l probably caure disruption of the aggregates, as they are formed, arid t h e rule will be fouiitl to hold only partially.
that any cweful study of the nature of reactions of this kind Iias been macle a i d would like to suggest it as a n interesting field for fnndniiiental research. To test this idea, several experiments were iiiade n-ith x-ariou. iubstaiices. If barium hydrate crystals, Ba(OH)2. 8H,O. are placed iii a drop of sulfate solution and observed with the iiiicroscope. it can be seen that tlie clieiiiical action is occurring oiily a t the surface of the crystals and that the coating o f hariuiii sulfate gradually increases iii thickness. Siiiiilarly, lead chloride crystals in an iodide zolution or lead iodide crystals in a cliromate solution show the reaction taking place a t tlie surface of the crystals ~ i t l i o udestruction t of the original particles. To confirm t h k idea that relative particle size may persist through the course of :I chemical reaction, harimii hydrate crystals were ground and screened iintil two i0-gr:im portions were obtained. One portioii was ccJiilposed of particles between 100 and 200 iiiesh and the yecoiid portion all passed the 200-mesh screen. Ea.ch portion was atltletl tri a solution containing 7 5 grams (equivalent to 5 per cent exce,ssj Sa28O4.10H20in 225 cc. of water, and the mixture was agitated a t room temperature on hottle rolls turning 92 r. p. 111. After various intervals settling tests were iiiade a r i d 10-cc. samples of the supernatant liquid were titrated with norinal hydrochloric acid to a phenolphthalein eiid point. The data are given in Table T?. Table \'I-Reaction TIMti
IXTERVAI.
'
,i HCI
-IOO
Summary
In the process of causticizing the nature of the final prc'cipit'ate can be greatly iiiodified and its settling rate c a n Iw changed a t least fifty fold, by changing the method of slaking and causticizing. Under standard coiiditions high-calciuiii limes froni n-itlely different sources give quite comparable results. The rate of reaction and the turbidity of the supernatatit, liquor are both changed by the m i l e factors that influence settling rate! so that' it is usually necessary to accept a ~ o n l I ) 1 ' l J mise in which the three factor. are all (misidered. Before an intelligent estimate c,an be iiiade of tlie 9ize of chemical equipment required for a given c,austiciziiig plant. all the factors affecting tlie settling arid reaction rate must be coiisidered and fixed. It has been showii that relative partide .
